XLV.] EEMAKKS ON THE LOGIC OF THE SCIENCES. 233 



remarked by Paracelsus, before tlie time of Bacon, theory Failure of 

 alone will lead to mere fantasy, and observation alone will ^^ . 



•^ Paracelsus. 



lead to mere empiricism. The best instance of the power- 



lessness of observation alone is perhaps the present state 

 of meteorology, or the science of the weather ; in which Mcteoro- 

 we have a vast mass of observations, and yet, for want of °^^' 

 knowing how to bring deductive theory to bear on them, 

 "we are almost as far as ever from knowing the laws and 

 causes on which the changes of the weather depend. The 

 powerlessness of theory alone is best shown by the utter 

 failure of all attempts to account for the facts of human 

 society and historical change by merely theoretical deduc- 

 tion from the laws of human nature. 



It is of course imjjossible that historical science can ever Historical 

 become mathematical in form ; nevertheless I think we are ^'^^*^"°® • 

 safe in asserting that the relation of theory to observation }ts niethod 



" •'IS tiie same 



is the same in the historical sciences as in the mathematico- as that of 

 physical ones, such as astronomy and the science of heat, science! 

 Certainty in the science of historical principles (of course 

 I do not mean mere historical facts) is attained only when 

 the laws generalized from the observed facts of history, 

 and the laws theoretically deduced from the general ten- 

 dencies of human nature, coincide so as to verify each 

 other. As an instance of this — not perhaps the best 

 instance that might be mentioned, but the most familiar — 

 I will mention the conclusion which all unprejudiced men 

 now accept, of the economical benefit of freedom of in- 

 dustry and exchange, or what is familiarly called free Free trade. 

 trade. This conclusion was first deduced by theory from 

 some of the commonest facts of human nature, and has 

 been amply verified by the experience of every state which 

 has had the wisdom to adopt it. 



It may be thought that mathematics is an exception to In what 

 the law of the necessity for verifying the results of the one ^athe- 

 method by those of the other ; — it may be thought that the matics 

 deductive results of pure or abstract mathematics need no fication. 

 verification. It is no doubt the fact that we take them as 

 true without demanding verification ; that is to say, when 

 we are satisfied of the accuracy of a mathematical calcula- 



